Intelligent task prediction and computation offloading based on mobile-edge cloud computing

Abstract Edge computing overcomes the high communication delay shortcoming of traditional cloud computing and provides computing services with high reliability and high bandwidth for mobile devices. At present, edge computing has become the forefront and hotspot of mobile-edge cloud computing (MEC) research. However, with the increasing requirements and services of mobile users, offloading strategy of simple edge computing is no longer applicable to MEC architecture. This paper puts forward a new intelligent computation offloading based MEC architecture in combination with artificial intelligence (AI) technology. According to the data size in computation task of mobile users and the performance features of edge computing nodes, a computation offloading and task migration algorithm based on task prediction is proposed. The computation task prediction based on LSTM algorithm, computation offloading strategy for mobile device based on task prediction, and task migration for edge cloud scheduling scheme are used to assist in optimizing the edge computing offloading model. Experiments show that our proposed architecture and algorithm can effectively reduce the total task delay and improve user experience with the amount of data and subtasks increase.